This study aimed to analyze critically the potential of Ipomoea nil'Scarlet O'Hara' for O(3) biomonitoring in the sub-tropics. Four field experiments (one in each season of 2006) were carried out in a location of the city of São Paulo mainly polluted by O(3). Each experiment started with 50 plants, and lasted 28 days. Sub-lots of five plants were taken at intervals between three or four days long. Groups of four plants were also exposed in closed chambers to filtered air or to 40, 50 or 80 ppb of O(3) for three consecutive hours a day for six days. The percentage of leaf injury (interveinal chloroses and necroses), the concentrations of ascorbic acid (AA) and the activity of superoxide dismutase (SOD) and peroxidases (POD) were determined in the 5th, 6th and 7th oldest leaves on the main stem of the plants taken in all experiments. Visible injury occurred in the plants from all experiments. Seasonality in the antioxidant responses observed in plants grown under field conditions was associated with meteorological variables and ozone concentrations five days before leaf analyses. The highest levels of antioxidants occurred during the spring. The percentage of leaf injury was explained (R(2) = 0.97, p < 0.01) by the reduction in the levels of AA and activity of POD five days before the leaf analyses and by the reduction in the levels of particulate matter, and enhancement of temperature and global radiation 10 days before this same day. Although I. nil may be employed for qualitative O(3) biomonitoring, its efficiency for quantitative biomonitoring in the sub-tropics may be compromised, depending on how intense the oxidative power of the environment is.